What problem does it solve? Building a governed semantic layer by hand requires inspecting physical schemas, inferring grain and relationships, and writing correct metric DDL — a slow, error-prone process. This Skill automates the derivation and maintenance of Exasol Semantic Views models so agents and analysts get deterministic, governed metrics. ## Core Features & Use Cases - Autonomous model derivation: Inspects tables, views, column comments, and historical query logs (EXA_DBA_AUDIT_SQL, profiling tables) to propose entities, relationships, facts, dimensions, and metrics in dependency order. - Full authoring lifecycle: Creates models, registers entities and grain proofs, authors additive/filtered/ratio/derived/window metrics, validates with VALIDATE_MODEL, and publishes governed semantic views. - Migration and fusion support: Imports Databricks Unity Catalog Metric Views and manages multi-representation fusion (F1 equivalents, F3 hot/cold partitions, F4 reconciliation, F5 identity mapping). - Use Case: Point the agent at a MART schema with order and customer tables; it inspects comments and workload history, derives a sales model with revenue and margin metrics, validates it, and publishes it for SQL, BI, and agent consumption. ## Quick Start Use the exasol-semantic-modeler skill to inspect the MART schema, derive a semantic model with entities, relationships, and metrics, validate it, and publish it as an Exasol Semantic Views model.
